Aplicação do modelo ARIMA-GARCH no processo de avaliação de empresas: um ensaio para o caso brasileiro

Abstract

This study seeks to propose an alternative methodology for valuation of Brazilian companies. This is done by incorporating econometric forecasting models for financial series into discounted cash flow model proposed by Damodaran (1997) in the prediction of cash flows, extending the analysis proposed by Canamary (2012) to the reality of 18 active companies in the Ibovespa index over the period between the first quarters of 2002 and 2015. By taking into account the effect of volatility in the evolution of cash flows, the current study uses the ARIMA-GARCH hybrid model. This study aimed to analyze the existence of significant differences between the distinct methodologies and determine which of these is more consistent with the estimates carried out from the first quarter of 2014 to the first quarter of 2015 through the metric of Mean Absolute Percentage Error (MAPE) and the test parametric t-student. Notwithstanding it was estimated the value of the company taking into account all the observations in the sample and the methodology which proved to be the most consistent in the process was identified. Compared to the traditional method of linear projection as emphasized by Varela and Vázquez (2009), the incorporation of the ARIMA-GARCH model to the technique of discounted cash flow (DCF) was effective in the evaluation process, presenting fewer inconsistencies in company value estimates and by considering the effect of volatility of these cash flows, helping to reduce the uncertainties inherent to the process.